Organic chemistry: Pharmaceutical diversification via palladium oxidative addition complexes

Article abstract of DOI:10.1126/science.aac6153

Palladium-catalyzed cross-coupling reactions have transformed the exploration of chemical space in the search for materials, medicines, chemical probes, and other functional molecules. However, cross-coupling of densely functionalized substrates remains a major challenge.We devised an alternative approach using stoichiometric quantities of palladium oxidative addition complexes (OACs) derived from drugs or drug-like aryl halides as substrates. In most cases, cross-coupling reactions using OACs proceed under milder conditions and with higher success than the analogous catalytic reactions. OACs exhibit remarkable stability, maintaining their reactivity after months of benchtop storage under ambient conditions.We demonstrated the utility of OACs in a variety of experiments including automated nanomole-scale couplings between an OAC derived from rivaroxaban and hundreds of diverse nucleophiles, as well as the late-stage derivatization of the natural product k252a.

Full text of DOI:10.1126/science.aac6153

RESEARCH
drug gefitinib to morpholine via the isolated
complex 1 (Fig. 1C), which was prepared by re-
acting gefitinib with the easily accessible Pd(0)
precursor [(tBuXPhos)Pd]2(COD) (26) at room
temperature. An extensive survey of ligands,
solvents, and bases revealed tBuXPhos with the
phosphazene base P2-Et (9–11) as the optimal
ligand-base combination for this coupling (27).
The desired product 2 was formed in 83% yield
in just 2 hours at room temperature by using
tBuXPhos with P2-Et in tetrahydrofuran (THF)
as solvent, whereas the analogous catalytic re-
action using 5 mole % (mol %) tBuXPhos Pd G3
failed completely. Using 1, reaction performance
was generally insensitive to the choice of solvent
[dimethyl sulfoxide (DMSO), dimethylformamide,
2-methyl-THF, toluene], and the coupling proceeded
effectively at concentrations as low as 0.01 M,
potentially facilitating automated (9, 28, 29) or
microfluidic synthesis (30, 31).
The stability of 1 was tested by storing it in a
vial on the benchtop under an atmosphere of air.
After 6 months, there was no observable loss
of purity or reactivity. Complex 1 also exhibited
good solution stability, providing comparable
yields whether a solution of the complex was
freshly prepared or aged for 18 hours prior to re-
action setup. In contrast to most catalytic C–N
cross-coupling reactions, reactions of 1 could also
be performed under an atmosphere of air with
only slightly lower efficiency (27). Similarly, the
OAC of the anticoagulant drug rivaroxaban (3)
could be isolated and coupled with morpholine
in 82% yield (Fig. 1D) to provide 4, whereas the
corresponding catalytic reaction gave a meager
2% yield. Analysis of 3 by single-crystal x-ray
diffraction confirmed the structure of this organo-
palladium complex derived from rivaroxaban,
a marketed drug (Fig. 1D).
Following these encouraging initial results, we
tested the generality of the OAC protocol on a
diverse set of drugs and drug-like aryl halides
(Fig. 2) whose complexity is represented by a
computed score (CQSAR) according to a recently
reported model (12). Several of these substrates
had been tested previously as part of a chem-
istry informer library (10) and provided con-
sistently low performance in C–N cross-coupling
reactions using state-of-the-art catalytic condi-
tions based on Pd, Cu, or Ni metallophotoredox
(32). The OACs derived from these halides, either
used as isolated solids or generated in situ, readily
coupled at room temperature with multiple
classes of amine nucleophiles to provide products
5 to 15 (Fig. 2). In all cases, the corresponding
catalytic reactions were less efficient and some-
times failed completely (compare 9 and 10, Fig. 2).
Although the use of isolated OACs improved the
likelihood that the first reaction attempt would
succeed, some substrates were still problematic,
such as 15 (Fig. 2) (27). In addition, whereas the
OAC method demonstrated a distinct advantage
over the catalytic method for most substrates of
moderate to high complexity (CQSAR = 2.5 to 3.5),
for a simple substrate such as 16 (CQSAR = 1.1)
there was no significant difference between the
two protocols. On occasion, we observed a minor
ORGANIC CHEMISTRY
Pharmaceutical diversification
via palladium oxidative
addition complexes
Mycah R. Uehling1,2, Ryan P. King2, Shane W. Krska1,
Tim Cernak1,3*, Stephen L. Buchwald2*
Palladium-catalyzed cross-coupling reactions have transformed the exploration of chemical
space in the search for materials, medicines, chemical probes, and other functional molecules.
However, cross-coupling of densely functionalized substrates remains a major challenge. We
devised an alternative approach using stoichiometric quantities of palladium oxidative addition
complexes (OACs) derived from drugs or drug-like aryl halides as substrates. In most cases,
cross-coupling reactions using OACs proceed under milder conditions and with higher success
than the analogous catalytic reactions. OACs exhibit remarkable stability, maintaining their
reactivity after months of benchtop storage under ambient conditions. We demonstrated the
utility of OACs in a variety of experiments including automated nanomole-scale couplings
between an OAC derived from rivaroxaban and hundreds of diverse nucleophiles, as well as the
late-stage derivatization of the natural product k252a.
alladium-catalyzed carbon-carbon (1) and
carbon-heteroatom (2, 3) bond-forming re-
actions have had a major impact on the
practice of organic synthesis and have be-
come mainstays of modern drug discovery
alyst regeneration steps. Because a successful
reaction outcome requires the efficient opera-
tion of each of these steps, interference from
a highly functionalized molecule with any one
step could, in principle, disrupt the entire desired
catalytic reaction. We reasoned that a Pd-mediated
cross-coupling reaction, where an equivalent of an
isolated oxidative addition complex (OAC; Fig. 1B)
(13) is used as the substrate, might improve the
likelihood that the reaction would succeed.
By design, such reactions initiate as close to
the product-forming step as possible (Fig. 1A,
highlighted region). Performing reactions with
the use of an OAC as the substrate obviates the
requirement for multiple rounds of catalyst turn-
over to form substantial amounts of product.
Although an OAC-mediated cross-coupling reac-
tion requires the use of one equivalent of pal-
ladium and ligand, the cost of these reagents
on a typical drug discovery scale (~$1 for a 25 mg–
scale reaction) is much less than that of a densely
functionalized drug-like substrate.
Palladium-mediated reactions are often used
instead of catalytic methods in settings where
first-pass reaction performance is considerably
more important than reagent cost, such as bio-
conjugation (14–16), peptide modification (17),
total synthesis (18–20), radiolabeling (21, 22),
PET imaging (23), and mechanistic inquiry
(24, 25). In previous studies from our group,
relatively simple organopalladium complexes
were used with high-complexity nucleophiles
such as peptides, proteins, and antibodies. With
this precedent, we wondered whether similar
organometallic complexes could be formed from
complex, drug-like aryl halides, and whether such
OACs would be stable, isolable, and practical for
use in a drug discovery context.
P
(4, 5). To date, most catalytic coupling methods
have been developed using relatively simple
model substrates (6). In contrast, the struc-
turally complex substrates typically encountered
in applied settings often contain functional groups
or substructures that can inhibit catalyst turnover
or participate in unproductive side reactions. For
example, many pharmaceutically relevant hetero-
cycles are competent ligands for transition metal
catalysts (7) and thus often function as competitive
catalyst inhibitors or displace activating ligands,
leading to inactive complexes (8). In practice, these
mechanistic vulnerabilities manifest as failed re-
actions; indeed, surveys of electronic notebooks
(9) and systematic evaluation of reactivity pat-
terns of collections of drug-like substrates (10, 11)
indicate that Pd-catalyzed cross-coupling reac-
tions of complex substrates fail frequently, with
failure rates above 50% for some chemistries.
With the trend toward increasing structural com-
plexity in synthetic drugs (12), new cross-coupling
approaches that exhibit greater reaction scope
and reliability will greatly facilitate the invention
of new medicines.
The generally accepted mechanism of Pd-
catalyzed cross-coupling reactions (Fig. 1A) in-
volves catalyst activation, oxidative addition,
nucleophile association or transmetallation,
deprotonation, reductive elimination, and cat-
1Merck & Co. Inc., Kenilworth, NJ 07033, USA. 2Department of
Chemistry, Massachusetts Institute of Technology, Cambridge,
MA 02139, USA. 3Department of Medicinal Chemistry, College of
Pharmacy, University of Michigan, Ann Arbor, MI 48109, USA.
*Corresponding author. Email: tcernak@umich.edu (T.C.);
sbuchwal@mit.edu (S.L.B.)
Indeed, organopalladium complexes derived
from drugs can be readily formed and isolated.
Initial efforts focused on coupling the anticancer
Uehling et al., Science 363, 405–408 (2019)
25 January 2019
1 of 4

RESEARCH
| REPORT
Fig. 1. Robust cross-coupling of
product formation
independent of
catalyst turnover
L
Pd
X
A
Aryl-Pd(II) complexes of drugs
B
drug-like molecules using Pd-based
oxidative addition complexes (OACs).
(A) Use of isolated OACs circumvents
many challenges associated with
achieving successful cross-coupling of
complex substrates. (B) OACs are stable,
readily handled, and economical at
discovery scale. (C and D) The complexes
(tBuXPhos)Pd•gefitinib (1) (C) and
(tBuXPhos)Pd•rivaroxaban (3) (D)
smoothly couple with morpholine
to give 2 and 4, respectively, whereas the
analogous catalytic reactions failed.
*Starting from (L)Pd•drug: P2-Et,
THF, room temperature (rt), 2 hours.
†Starting from aryl halide: 5 mol %
tBuXPhos Pd G3, P2-Et, THF, rt, 18 hours.
X
OAC
oxidative addition
complex (OAC)
oxidative
addition
nucleophile
association
NuH
L
Pd
drug
LPd0
X
NuH+
Pd L
X
catalyst
activation
HX
(L)Pdn
deprotonation
stable to moisture and storage in air
~ $1 per reaction on 25 mg scale
catalyst
reductive
elimination
regeneration
Nu
improved coupling performance with
complex substrates
F
F
C
O
tBuXPhos
HN
O
HN
N
Pd
O
HN
N
N
O
Cl
N
N
O
O
N
conditions
2
MeO
1
gefitinib(Pd)CltBuXPhos
MeO
N
cat†
*
OAC
83% 0%
O
D
O
NH
O
HN
O
N
S
conditions
O
N
N
4
cat†
2%
*
OAC
80%
3
O
O
rivaroxaban(Pd)CltBuXPhos (X-ray)
complication that substrates or products underwent
apparent decomposition reactions that were at-
tributed to the L•Pd(0) species generated from
the OAC during the product release step. In such
cases (5–7, 11, 12, Fig. 2), this decomposition
could be mitigated by adding a sacrificial aryl
halide such as 4-bromoanisole to quench the
L•Pd(0) species (27). Overall, the results in Fig. 2
indicate that this cross-coupling protocol can be
used with a high rate of success on a variety of
complex aryl halides and diverse amine coupling
partners to obtain sufficient product to test bio-
logical and physical properties.
G3 and related Pd(II) complexes enhances the
applicability of the OAC strategy, because many
air- and moisture-stable Pd complexes are com-
mercially available. However, this approach re-
quires the use of an excess of base to form Pd(0).
Additionally, one equivalent of a carbazole by-
product is formed during the activation of the
Pd complex. Both of these attributes can compli-
cate purification of the coupling product. In ad-
dition, the in situ formation of OACs is challenging
with aryl halides that have low solubility in or-
ganic solvents, potentially limiting the applica-
tion of this approach in automated or flow
synthesis. In contrast, every OAC we generated
from [(tBuXPhos)Pd]2(COD) was soluble under
the reaction conditions described. As an example,
rivaroxaban (precursor to 3) wasinsoluble in DMSO,
whereas the corresponding isolated OAC 3 rapidly
and completely dissolved in the same solvent.
The high solubility of OACs in organic solvents
makes them attractive substrates for use in auto-
mated parallel synthesis or microfluidic synthesis.
This was demonstrated by synthesizing more
than 200 analogs of 3 in parallel via nanoscale
synthesis (Fig. 3). Cross-coupling reactions of 3
with 384 diverse C-, N-, O-, and S-nucleophiles
were performed using a nanoscale robotic plat-
form (9), and the desired products were observed
in 206 of the 384 distinct reactions (>10% con-
version observed in 139 instances). For compar-
ison, in the analogous catalytic cross-coupling
of the corresponding aryl halide rivaroxaban (17)
performed with the same 384 nucleophiles, prod-
ucts were observed in 39 instances (>10% con-
version observed in one reaction; Fig. 3).
During a medicinal chemistry campaign, it is
desirable to have modular control over target
structures. It was therefore interesting that from
a single set of reaction conditions, OAC 3 had
relatively consistent coupling performance across
a diverse set of nucleophiles: 49% of the amines
(n = 172), 37% of the alcohols (n = 30), 36% of the
thiols (n = 14), and 42% of the boronic acids or
boronic esters (n = 12) showed >10% conver-
sion to product. Although more than half of the
nucleophiles in every class did not deliver ap-
preciable amounts of product, the corresponding
catalytic method showed >10% conversion to
product for just one nucleophile, likely owing
in part to the fragile nature of the oxazolidinone
ring of 17. Selected C–N cross-coupling examples
from the nanoscale synthesis experiment were
repeated at 20 mg scale, and products 18 to 24
were formed in 41 to 86% yield. Except for the
reaction to form product 25 from a sterically
hindered amine, the Pd-based couplings using
OAC 3 delivered more product than the analo-
gous catalytic reactions (Fig. 3).
We also investigated the use of a stoichiometric
t
quantity of BuXPhos Pd G3 for these trans-
formations. We hypothesized that this approach
would form, in situ, the same OACs generated
from [(tBuXPhos)Pd]2(COD). We observed the
rapid formation of OAC 3 after mixing rivar-
oxaban, P2-Et, and 100 mol % of tBuXPhos Pd G3
as judged by 31P nuclear magnetic resonance.
In the presence of morpholine and excess P2-Et,
however, formation of 3 could not be detected
because cross-coupling to produce 4 was com-
plete before we were able to analyze the first
reaction time point (<2 min). After mixing aryl
t
halides and amines with 100 mol % BuXPhos
To further demonstrate the breadth of cross-
coupling reactions that can be executed from
preformed OACs, we subjected 3 to a variety of
carbon-carbon and carbon-heteroatom bond-
forming reaction conditions to generate a rich
array of products from this complex drug core
Pd G3 and P2-Et, anilines 2, 4, 7, 12, and 14
were produced in 55%, 88%, 68%, 68%, and 43%
yield, respectively. This is comparable to the
yields observed using isolated OACs (Figs. 1 and
2). In situ formation of OACs using tBuXPhos Pd
Uehling et al., Science 363, 405–408 (2019)
25 January 2019
2 of 4

RESEARCH
| REPORT
L
Pd
X
R
N
R'
legend
yield
61-100%
41-60%
11-40%
0-10%
NR(R')H
LPd(0)
THF
X
CQSAR
ID
P2-Et, THF,
rt, 1-3 h
OAC cat.
Drugs or drug
precursors
oxidative addition complex (OAC)
isolated or prepared in situ
F
F
F
O
O
O
O
O
O
O
O
O
O
O
MeO2C
N
N
N
N
MeO
MeO
MeO
tBuO
N
BnO
O
tBuO
N
tBuO
N
N
N
N
N
N
O
HN
N
O
O
O
*
*
6
*
7
5
3.3
3.2
3.2
8
3.4
O
OEt
67% 16%
48% 10%
70% 24%
38% 5%
Me
Me
H
N
Me
Me
O
N
O
O
N
N
O
S
Me
O
S
O
O
N
O
O
O
O
Me
N
F
O
O
O
N
F3C
N
N
Me
( )6
N
N
H
O
,
10†
2.5
11
2.6
2.6
*
*
12
9
3.0
39% 0%
42% 0%
71% 28%
45% 30%
Me
Me
CN
H
N
CO2Me
N
O
N
Me
Me
O
Me
H
Me
O
O
N
N
N
O
N
N
OEt
N
O
N
H
N
N
NC
O
tBuO
NH2
2.8
13
3.5
14
2.7
15
16
1.1
56% 15%
52% 33%
0% 0%
88% 80%
Fig. 2. Complex aryl halide couplings. *1.0 equiv 4-bromoanisole added. †1.0 equiv amine and 1.0 equiv P2-Et. ⊥Oxidative addition performed in situ. Yields
from OACs are the average of two runs and were in good agreement, except for compound 12 where some variability was observed. See (27) for details.
X
Nanoscale synthesis using Pd catalyst
Nanoscale synthesis using Pd OAC
17
5 mol%
H
N
S
tBuXPhos Pd G3
P2-Et
legend
yield
61-100%
41-60%
11-40%
0-10%
3
75
25
0
75
25
0
O
P2-Et
CQSAR
ID
OAC cat.
O
N
O
3
X = Pd(Cl)tBuXPhos
17 X = Cl
N
Ar
Ar
N
Ar
N
NH
Ar
N
O
O
N
OMe
N
O
N
N
Me
N
OEt
3.1
Me
20
84% 0%
18
3.2
19
3.1
21
3.1
82% 15%
68% 0%
41% 0%
Ar
N
H
N
Ar
Ar
Me
Me
Ar
N
N
N
H
Me
O
22
3.0
0%
23
3.0
24
3.1
25
3.1
63% 10%
78% 0%
0% 0%
86%
Fig. 3. Nanoscale coupling of rivaroxaban with 384 diverse nucleophiles. Reactions were performed at 0.05 mg scale on a robotic platform to compare
t
the performance of 17 (conditions: 5 mol % BuXPhos Pd G3, P2-Et, DMSO, rt, 18 hours) to the performance of 3 (P2-Et, DMSO, rt, 2 hours) in coupling
reactions with 384 diverse nucleophiles. Select examples from nanoscale synthesis (18–24) were repeated at 20 mg scale. See (27) for details.
Uehling et al., Science 363, 405–408 (2019)
25 January 2019
3 of 4

RESEARCH
| REPORT
Fig. 4. Late-stage derivatization of
the drug rivaroxaban and the
natural product k252a. (A) Complex 3
readily participates in diverse
coupling transformations.
(B) Complex 33 derived from
natural product k252a is readily
trifluoromethylthiolated. In
Me
B
A
O
O
O
O
B
O
HN
B
N
ZnBr
2.9
S
Me
O
Me
2.7
O
Ar Me
Ar
Cl
3
Pd
tBuXPhos
26
29
N
H
N
51% 0%
47% 5%
O
CO
O
all cases, analogous catalytic
reactions gave <5% yield. See (27)
for details.
CF3SAg
Zn(CN)2
Ar SCF3
Ar CN
O
27
2.9
30
2.9
32
3.0
Ar
N
78% <5%
95% 5%
82% 0%
O
CO2Me
NHBoc
HS
31
Ar
NHBoc
legend
yield
S
NHBoc
3.4
61-100%
41-60%
11-40%
0-10%
CQSAR
ID
OAC cat.
CO2Me
NHBoc
28
3.1
Ar
65% 0%
90% 0%
H
N
H
N
B
O
O
Br
Pd
SCF3
4.3
tBuXPhos
AgSCF3
N
N
N
N
O
O
Me
Me
34
33
MeO2C
MeO2C
82% 0%
HO
HO
(Fig. 4A). Reactions of 3 with carbon-based
nucleophiles such as cyclopropyl zinc bromide
(26), N-Boc-propargyl amine (28), methyl bor-
oxine (29), and zinc cyanide (30) gave the cor-
responding cross-coupling products in good to
excellent yields. Compound 3 also reacted ef-
ficiently with S-based nucleophiles such as silver
trifluoromethyl sulfide (27) and N-Boc-cysteine
(31). The potential for OACs to participate in
carbonylative amination was demonstrated by
the formation of amide 32 in 82% yield from the
reaction of 3 with morpholine under an atmo-
sphere of carbon monoxide. Finally, complex 33
(Fig. 4B), prepared via selective bromination of
the natural product k252a followed by reaction
with (tBuXPhosPd)2(COD), underwent cross-
coupling smoothly with silver trifluoromethyl
sulfide to provide 34 (CQSAR = 4.3) in 82% yield.
As with all other substrates examined in this
study, the reaction based on use of OACs suc-
ceeded while the corresponding catalytic reac-
tions were much lower-yielding or failed altogether.
Our data suggest that the use of OACs will
increase the success rate of cross-coupling re-
actions and will accelerate access to molecules
of high complexity. For most of the transforma-
tions described, successful catalytic conditions
could likely be identified given sufficient time
and resources spent on optimization. However, it
is precisely time and resources that are in short
supply during drug discovery. In practical terms,
synthetic methods that increase the speed and
efficiency with which new molecules can be ac-
cessed for primary testing will have a positive
impact on pharmaceutical development overall.
Beyond this, methods such as these that allow
access to high-complexity chemical space offer
the potential to change the types of molecules that
medicinal chemists target for biological testing.
In this context, we believe that the use of preformed
organometallic complexes, such as the Pd-based
OACs described here, holds tremendous promise
for translating catalytic methods of limited scope
into high-value chemical transformations that
operate on complex molecules, opening new syn-
thetic possibilities for molecule inventors.
22. T. L. Andersen et al., J. Am. Chem. Soc. 137, 1548–1555
(2015).
23. E. Lee et al., Science 334, 639–642 (2011).
24. H. M.-F. Viart, A. Bachmann, W. Kayitare, R. Sarpong, J. Am.
Chem. Soc. 139, 1325–1329 (2017).
25. D. M. Ferguson, J. R. Bour, A. J. Canty, J. W. Kampf,
M. S. Sanford, J. Am. Chem. Soc. 139, 11662–11665
(2017).
26. H. G. Lee, P. J. Milner, M. T. Colvin, L. Andreas, S. L. Buchwald,
Inorg. Chim. Acta 422, 188–192 (2014).
27. See supplementary materials.
28. J. Li et al., Science 347, 1221–1226 (2015).
29. A. G. Godfrey, T. Masquelin, H. Hemmerle, Drug Discov. Today
18, 795–802 (2013).
30. K. F. Jensen, AIChE J. 63, 858–867 (2017).
31. D. Perera et al., Science 359, 429–434 (2018).
32. E. B. Corcoran et al., Science 353, 279–283 (2016).
REFERENCES AND NOTES
1. A. de Meijere, F. Diederich, Metal-Catalyzed Cross-Coupling
Reactions (Wiley-VCH, ed. 2, 2004).
2. P. Ruiz-Castillo, S. L. Buchwald, Chem. Rev. 116, 12564–12649
(2016).
3. J. F. Hartwig, Acc. Chem. Res. 41, 1534–1544 (2008).
4. N. Schneider, D. M. Lowe, R. A. Sayle, M. A. Tarselli,
G. A. Landrum, J. Med. Chem. 59, 4385–4402 (2016).
5. D. G. Brown, J. Boström, J. Med. Chem. 59, 4443–4458
(2016).
6. A. Nadin, C. Hattotuwagama, I. Churcher, Angew. Chem. Int. Ed.
51, 1114–1122 (2012).
7. E. C. Hansen et al., Nat. Chem. 8, 1126–1130 (2016).
8. M. A. Düfert, K. L. Billingsley, S. L. Buchwald, J. Am. Chem. Soc.
135, 12877–12885 (2013).
9. A. Buitrago Santanilla et al., Science 347, 49–53 (2015).
10. P. S. Kutchukian et al., Chem. Sci. 7, 2604–2613 (2016).
11. N. J. Gesmundo et al., Nature 557, 228–232 (2018).
12. R. P. Sheridan et al., J. Chem. Inf. Model. 54, 1604–1616
(2014).
13. B. T. Ingoglia, S. L. Buchwald, Org. Lett. 19, 2853–2856 (2017).
14. E. V. Vinogradova, C. Zhang, A. M. Spokoyny, B. L. Pentelute,
S. L. Buchwald, Nature 526, 687–691 (2015).
15. A. J. Rojas, B. L. Pentelute, S. L. Buchwald, Org. Lett. 19,
4263–4266 (2017).
ACKNOWLEDGMENTS
We thank R. Liu, A. Thomas, Y.-M. Wang, N. White, and C. Nguyen for
advice on the preparation of this manuscript; C. Tsay for x-ray
crystallographic analysis of 3; H.-Y. Kim, A. Beard, and P. Chen for
characterization support; and an anonymous reviewer for discussion
focused around the use of stoichiometric palladium. Funding:
Supported by Merck Sharp & Dohme Corp., a subsidiary of Merck
& Co. Inc. Author contributions: S.L.B. came up with the
original idea; T.C., M.R.U., S.L.B., and S.W.K. conceptualized the
strategy; M.R.U., T.C., and S.L.B. designed the experiments;
M.R.U., T.C., and R.P.K. performed the experiments; and M.R.U.,
T.C., S.W.K., and S.L.B. wrote the manuscript. Competing interests:
MIT holds patents on the ligands and precatalysts used in this
work, from which S.L.B. and former co-workers receive royalty
payments. Data and materials availability: The x-ray data for
complex 3 are freely available from the Cambridge Crystallographic
Data Centre under CCDC 1863599. All other data are available in
the supplementary materials.
16. A. J. Rojas et al., Chem. Sci. 8, 4257–4263 (2017).
17. T. D. Kondasinghe, H. Y. Saraha, S. B. Odeesho, J. L. Stockdill,
Org. Biomol. Chem. 15, 2914–2918 (2017).
18. P. S. Baran, C. A. Guerrero, E. J. Corey, J. Am. Chem. Soc. 125,
5628–5629 (2003).
19. D. L. Boger, S. R. Duff, J. S. Panek, M. Yasuda, J. Org. Chem.
50, 5782–5789 (1985).
20. R. L. Simmons, R. T. Yu, A. G. Myers, J. Am. Chem. Soc. 133,
15870–15873 (2011).
SUPPLEMENTARY MATERIALS
www.sciencemag.org/content/363/6425/405/suppl/DC1
Materials and Methods
Figs. S1 to S17
Tables S1 to S4
NMR Spectra
References (33–37)
21. H. Yang, P. G. Dormer, N. R. Rivera, A. J. Hoover, Angew. Chem.
Int. Ed. 57, 1883–1887 (2018).
15 May 2018; accepted 18 December 2018
10.1126/science.aac6153
Uehling et al., Science 363, 405–408 (2019)
25 January 2019
4 of 4

Pharmaceutical diversification via palladium oxidative addition complexes
Mycah R. Uehling, Ryan P. King, Shane W. Krska, Tim Cernak and Stephen L. Buchwald
Science 363 (6425), 405-408.
DOI: 10.1126/science.aac6153
Embedding palladium ahead of time
Palladium-catalyzed cross-coupling is one of the most widely applied reaction classes in pharmaceutical research.
The metal is adept at connecting aromatic rings to one another or to nitrogen centers. However, functional complexity
can obstruct the reaction, necessitating laborious ligand optimization. Uehling et al. mitigated this problem by isolating
the stable product of palladium's reaction with a complex aryl halide ahead of time. Subjecting these compounds to
downstream coupling reactions substantially improved yields.
Science, this issue p. 405
ARTICLE TOOLS
http://science.sciencemag.org/content/363/6425/405
SUPPLEMENTARY
MATERIALS
http://science.sciencemag.org/content/suppl/2019/01/23/363.6425.405.DC1
REFERENCES
This article cites 36 articles, 5 of which you can access for free
http://science.sciencemag.org/content/363/6425/405#BIBL
PERMISSIONS
http://www.sciencemag.org/help/reprints-and-permissions
Use of this article is subject to the Terms of Service
Science (print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement of
Science, 1200 New York Avenue NW, Washington, DC 20005. 2017 © The Authors, some rights reserved; exclusive
licensee American Association for the Advancement of Science. No claim to original U.S. Government Works. The title
Science is a registered trademark of AAAS.